Stigmergy, communicating through the environment, plays a major role in organizing ant colony behaviour. Most of that stigmeric communication is done through scents, and most of the examples of stigmeric communication I mention in Stigmergy and the World-Wide Web are examples where the scent is in the static physical environment, that is, in the mud in the walls of the nest and on the ground on an ant trail. Now researchers at The Gordon Lab at Stanford have discovered that ants performing differing tasks smell differently, and that the ant behaviour changes in the face of collections of ants with particular scents. This is just another example of stigmergy, even though the scent is on the other ants it is not direct communication from ant to ant, as only a group of ants with the same smell will trigger a new behaviour in the ants.
It makes sense from an evolutionary point of view, that is, if you have a successful strategy, keep re-using it for maximum effect. In other words, even in nature, if you have a hammer, everything looks like a nail. There is more detail in the Stanford press release "Work stinks": It's more than just a slogan among ants, researchers find [via context weblog]